Competition is often proposed to drive niche segregation along multiple axes in speciose communities. Understanding spatial partitioning of foraging areas is particularly important in species that are constrained to a central place. Contemporary habitat partitioning is sometimes hypothesised to be evidence of competition shaping coevolution, citing a potential ‘ghost of competition past’ (Connell 1980). In this talk, I will present a natural experiment examining variation in habitat preferences of congeneric Southern Ocean predators in sympatry and allopatry, to show that ghost-hunting is much harder than it first appears. I will also discuss some upcoming work at the interface of seabird movement ecology and conservation planning.